The aims of this paper are to review previously published palaeovegetation and independent palaeoclimatic datasets together with new results we present from dynamic vegetation model simulations and modern pollen rain studies to: (i) determine the responses of Amazonian ecosystems to changes in temperature, precipitation and atmospheric CO2 concentrations that occurred since the Last Glacial Maximum (LGM), ca. 21 000 years ago; and (ii) use this long–term perspective to predict the likely vegetation responses to future climate change. Amazonia remained predominantly forested at the LGM, although the combination of reduced temperatures, precipitation and atmospheric CO2 concentrations resulted in forests structurally and floristically quite different from those of today. Cold–adapted Andean taxa mixed with rainforest taxa in central areas, while dry forest species and lianas probably became important in the more seasonal southern Amazon forests and savannahs expanded at forest–savannah ecotones. Net primary productivity (NPP) and canopy density were significantly lower than today. Evergreen rainforest distribution and NPP increased during the glacial—Holocene transition owing to ameliorating climatic and CO2 conditions. However, reduced precipitation in the Early–Mid–Holocene (ca. 8000–3600 years ago) caused widespread, frequent fires in seasonal southern Amazonia, causing increased abundance of drought–tolerant dry forest taxa and savannahs in ecotonal areas. Rainforests expanded once more in the Late Holocene owing to increased precipitation caused by greater austral summer insolation, although some of this forest expansion (e.g. in parts of the Bolivian Beni) is clearly caused by palaeo Indian landscape modification. The plant communities that existed during the Early–Mid–Holocene may provide insights into the kinds of vegetation response expected from similar increases in temperature and aridity predicted for the twenty–first century. We infer that ecotonal areas near the margins of the Amazon Basin are liable to be most sensitive to future environmental change and should therefore be targeted with conservation strategies that allow ‘natural’ species movements and plant community re–assortments to occur.